Package org.ddogleg.optimization
package org.ddogleg.optimization
-
ClassDescriptionLambda intended to be used to adjust the value of an array.Configuration for
GaussNewtonBase_F64.Configuration for built inLoss Functions.General configuration for unconstrained non-linear least squares solvers.Convergence paramters forUnconstrainedMinimizationandUnconstrainedLeastSquares.Used to validate an algebraic Jacobian numerically.Factory for creating differentLossFunctionandLossFunctionGradient.Functions for creating numerical derivativesCreates optimization algorithms using easy to use interfaces.Factory for sparse optimization algorithms.Base class for Gauss-Newton based approaches for unconstrained optimization.Optimization modeInterface for iterative optimization classes.LeastSquaresSwitcher<S extends DMatrix>Class that lets you easily switch between differentGaussNewtonBase_F64solvers and matrix formats.Abstracts writing to the Jacobian so thatLeastSquaresSwitcher.ProcessJacobianOutdoesn't need to know the formatHigh level interface for implementing the Jacobian.Line search for nonlinear optimization.OptimizationDerivative<State>Interface for computing the gradient of a set of functions given a set of model parameters.This message is thrown if something bad happens while optimizing that would be the results invalidUnconstrainedLeastSquares<S extends DMatrix>Non-linear least squares problems have a special structure which can be taken advantage of for optimization.Common base for implementations ofUnconstrainedLeastSquaresthat differ in the JacobianFunctionUnconstrainedLeastSquaresSchur<S extends DMatrix>A variant onUnconstrainedLeastSquaresfor solving large scale systems which can be simplified using the Schur Complement.Optimization algorithm which seeks to minimize F(X) ∈ ℜ and X ∈ ℜNPerforms common optimization tasks.